analiza aiqbench
This research introduces a novel framework that enhances Large Language Model-based time-series anomaly detection by explicitly incorporating frequency-domain evidence derived from Fast Fourier Transform (FFT). By preserving both indexed observations and spectral structure, the method effectively captures recurring temporal patterns like shifted periodicity that traditional time-domain approaches often miss. The study demonstrates that augmenting LLM inputs with this compact frequency data significantly improves zero-shot detection performance across multiple state-of-the-art models.
Structured Frequency-Domain Evidence for LLM-Based Time-Series Anomaly Detection
arXiv:2608.24113v1 Announce Type: cross Abstract: Time-series anomalies can appear not only as pointwise deviations but also as changes in recurring temporal structure, such as shifted periodicity or localized oscillatory fluctuations. However, existing LLM-based time-series anomaly detection methods mainly expose time-domain evidence through indexed values, plots, or de-seasonalized representations, leaving spectral structure implicit. We propose an evidence-augmented zero-shot TSAD framework that preserves indexed de-seasonalized observations while adding compact frequency-domain evidence computed with the Fast Fourier Transform (FFT). The evidence is constructed at two resolutions: global frequency-domain evidence summarizes sequence-level periodic context, while local frequency-domain evidence captures time-localized spectral departures. Experiments on AnomLLM with InternVL2-LLaMA3-76B, Qwen2.5-VL-72B-Instruct, Gemini-2.5-Flash, and GPT-4o, together with evaluation on the TSB-AD-U subset, show that explicit frequency-domain evidence improves LLM-based TSAD baselines. These results suggest that frequency-domain evidence can complement indexed and de-seasonalized time-domain inputs for zero-shot LLM-based TSAD.
Abstract:Time-series anomalies can appear not only as pointwise deviations but also as changes in recurring temporal structure, such as shifted periodicity or localized oscillatory fluctuations. However, existing LLM-based time-series anomaly detection methods mainly expose time-domain evidence through indexed values, plots, or de-seasonalized representations, leaving spectral structure implicit. We propose an evidence-augmented zero-shot TSAD framework that preserves indexed de-seasonalized observations while adding compact frequency-domain evidence computed with the Fast Fourier Transform (FFT). The evidence is constructed at two resolutions: global frequency-domain evidence summarizes sequence-level periodic context, while local frequency-domain evidence captures time-localized spectral departures. Experiments on AnomLLM with InternVL2-LLaMA3-76B, Qwen2.5-VL-72B-Instruct, Gemini-2.5-Flash, and GPT-4o, together with evaluation on the TSB-AD-U subset, show that explicit frequency-domain evidence improves LLM-based TSAD baselines. These results suggest that frequency-domain evidence can complement indexed and de-seasonalized time-domain inputs for zero-shot LLM-based TSAD.
Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
arXiv:2608.24113 [cs.LG]
(or arXiv:2608.24113v1 [cs.LG] for this version)
https://doi.org/10.48550/arXiv.2608.24113 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Text integral sub licența deschisă a publicației · 232 cuvinte
Subiecte conexe
Vertical Advantage: Transforming Industries with Lakebase and Agentic AI
Databricks · 27 aug. 2026
7 Steps to Mastering Time Series Analysis with Python
KDnuggets · 4 iun. 2026
Beyond Uniform Tokens: Adaptive Compression for Time Series Language Models
arXiv · 12 iun. 2026
Game Design Creativity evals
LessWrong · 12 oct. 2026